Intermittent or Cyclic Faults: Diagnostic Failures in Psychiatry

A few years ago my car kept breaking down due to a flat battery. The fault was intermittent. It would be ok for a few days, then the battery would be flat again. First thought was either a dud battery or the alternator (which most people will know, is what charges your battery as you’re driving). It turned out to be neither of these. I then realise it was probably an intermittent electrical fault.

I took my car to several garages who all tested the electrics. They all told me that there was nothing wrong with the electrics. This was astounding. They weren’t listening to me. I would respond something like “Yes, but every few days the battery is flat, and it’s not the alternator or the battery, so surely it must be the electrics”. They would then say, “Well all our tests have come back clear”.

Not a Bug!


This is when my blood usually boils. I was trained in computer science, which included training in software engineering / programming. Debugging systems is part of software engineering.

Debug Definition

You have to do detailed diagnostics to find the problem (the bug or error). Often the worst problems to find are what are known as intermittent bugs, that is bugs that only occur occasionally. A lot of people, including some software engineers, will say that there isn’t a bug if they can’t find it. In this instance, they are completely overruling the experience of the person reporting the issue. Sometimes it is true that the person reporting the issue is mistaken, and the problem lies elsewhere, however you mustn’t say ‘not a bug’ without doing more detailed analysis and checking with the person who reported the issue.

Jobs like psychiatry, psychology, car mechanic, software engineer, hardware engineer, all need good diagnostic skills. Unfortunately these skills are rarely taught well, even on most computer science degrees! This shocks me because it is one of the most important skills of these jobs. Trial and error, and blagging it doesn’t cut it. This is because a lot of people have cognitive errors in relation to diagnostics. The biggest cognitive error is:

I can’t see it, so it doesn’t exist

When I’ve seen psychiatrists, and even GPs, I’ve found their diagnostic skills to be poor. For example if someone has insomnia, the first thing to understand is that insomnia has multiple causes. The second is that someone could have more than one of these multiple causes occurring in their situation, so ruling just one out might not fix the problem. Questions are important when diagnosing a problem. For example “Is it worse at certain times, or is the problem consistent every night?”. Most GPs fail to ask enough questions or do enough tests to diagnose the problem accurately.  The same is true in psychiatry.

Accurate Measures (e.g. Moodcharts)

I almost certainly have undiagnosed bipolar disorder. I have periods where I feel amazing, and other times where I completely crash. Even if my life is going well, the storms will still come, meaning it seems more neurological than down to external events. When I’ve made appointments with my GP to see a psychiatrist, which I’ve done over two decades, I only make the appointments when I’m well! So they never see me when I’m unwell. I’ve explained this to every psychiatrist in the past several years, and yet they fail to listen to this extremely important point. If they only see me when I’m well, then they think I’m ok, just like the car mechanic with my electrical fault (which turned out to be a problem with the earthing points on the car). The psychiatrist gave me a mood chart last time, and asked me to do it for 2 months. I explained that my cycle is wider than that, i.e. I warned her that the mood chart would have to go on for longer to accurately measure my moods. She didn’t listen (of course… psychiatrists rarely listen to their patients!). So of course I do the mood chart. It comes back normal (i.e. no extreme highs or lows). I see her again, and she says “You seem ok, this is normal”. I explain again that it needs to be done for longer… no response… oh dear. My life is sometimes at risk due to being suicidal, and yet… no response… oh my.

A couple of months later I have a major crash where I end up in A&E wanting to kill myself. Does this smack of negligence to you? It does to me. All because most psychiatrists have poor diagnostic skills (as do a lot of the other professions I mentioned).

Improved Training

Diagnostic skills should be taught in a lot more detail, and the common cognitive errors that can stop effective diagnostic skills need to be looked at. Software engineers, like myself, need to get together with other people from other professions who have excellent diagnostic skills, and we need to come up with a course. We then need to train psychiatrists and doctors with these skills. We should probably teach people from other professions too, however this is a mental health blog, so I’m more interested in the medical community learning better diagnostic skills.

What are your experiences or thoughts on diagnostic skills? Have you seen professionals fail badly with these skills? Do you possess good diagnostic skills? Let me know in the comments.

Copyright MEN HEAL 2016

Leave a Reply

Your email address will not be published. Required fields are marked *